HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Search for Related Content Social Bookmarking                      What's this?

Breast

Vilar V.; Goldman S.; Demarchi G.; Abud T.; Szejnfeld J. UNIFESP, Sao Paulo, Brazil

Address correspondence to V. Vilar (vanvilar76{at}gmail.com)

Background: Diabetic mastopathy (MD), a rare entity is an uncommon linfocitic chronic mastitis that occurs typically in middle age women type I insulin-dependent diabetes mellitus (DM). We plan to describe the most important clinical, radiological, and anatomopathologic (AP) findings of this disease which, most of the time, overlap the ones of breast cancer.

Key Issues: Mammography usually demonstrates extremely dense breasts and US shows an irregular mass with posterior acoustic shade (fibrosis). MR shows breast asymmetry, high and diffuse signal of the skin and parenchyma, heterogeneous mass, with dynamic progressive enhancement curve within the spiculate areas.

Format: Didactic presentation: A-Etiopathogeny. B-Physiopathology. C-Clinical appearance. D-Imaging findings: MMG, US and MR. E-Anatomopathologic findings. F-Case report. G-Differential diagnosis.

Teaching Points: MD is an uncommon benign disease; however, it can present clinical, imaging, and AP findings that overlap those from breast invasive or inflammatory carcinoma. We hope the viewer will learn first-hand about this disease and how its findings relate to those of more common breast pathologies.

E002. Bilateral Diabetic Fibrous Mastopathy of the Breast: Radiologic Findings with Pathologic Correlation

Woo O.; Kim J.; Yong H.; Kim A.; Seol H.; Kang E. Korea University Guro Hospital, Seoul, South Korea

Address correspondence to O. Woo (wokhee{at}unitel.co.kr)

Objective: Diabetic fibrous mastopathy is an unusual cause of benign breast masses that is characteristically seen in patients with long standing, insulin-dependent diabetics. The purpose of this study is to know radiologic findings, whenever possible mammography, ultrasound and/or MR imaging and pathologic correlation of bilateral diabetic fibrous mastopathy (BDFM) and to understand radiologic points which are helpful in differentiating it from bilateral breast carcinoma.

Materials and Methods: We retrospectively reviewed the radiologic findings of eight patients with pathologically proven BDFM, which were confirmed by ultrasound-guided core biopsy (n = 5) or by surgical excision (n = 3). The patients included all women, mean age 38 years (range, 20-56 years). Patients underwent mammography and breast ultrasound (n = 8), and magnetic resonance imaging (MRI) (n = 3).

Results: Mammography showed asymmetric increased density with ill-defined margins in all lesions. Breast US showed a heterogeneously hypoechoic mass with ill-defined or spiculated margins (n = 8) and marked posterior shadowing (n = 6). All lesions interrogated with color flow US showed no Doppler signal. MR imaging revealed a low signal intensity on T2WI, suggestive of fibrosis (n = 1), nonspecific stromal enhancement (n = 1) and gradual enhancement on postcontrast T1WI (n = 1).

Conclusion: BDFM mimics bilateral breast cancer, especially bilateral lobular carcinoma. BDFM shows no Doppler signal on color flow US and no definite evidence of strong enhancement on MR imaging, in contrast with bilateral breast cancer.

E003. Stromal Fibrous Mastopathy: Clinical and Radiological Aspects

Vilar V.; Goldman S.; Demarchi G.; Abud T.; Szejnfeld J. UNIFESP, Sao Paulo, Brazil

Address correspondence to V. Vilar (vanvilar76{at}gmail.com)

Background: Stromal fibrous mastopathy (FM) is a benign pathologic entity characterized by abundant proliferation of fibrous connective stromal tissue.

Key Issues: We plan to describe the clinical appearance and imaging findings at mammogram (MMG), ultrasound (US) and magnetic resonance (MR) which, most of the time, overlap the ones of breast cancer. FM has nonspecific clinical and imaging presentations. At palpation it presents as a hard mass with little mobility, cutaneous thickening or retraction. MMG demonstrates dense breasts and US shows hypoechogenic irregular areas with posterior acoustic shade.

Format: Didactic presentation: Etiopathogeny. Clinical appearance. Imaging findings: MMG, US and MR. Case reports. Differential diagnosis.

Teaching Points: It is important to recognize its main aspects so that a correct histopathologic diagnosis of an initially suspected area is not considered incompatible, since its appearance may overlap that of breast cancer.

E004. MR Imaging of Fibrocystic Breast

Chen J.¹; Su M.²; Nalcioglu O.² 1. China Medical University Hospital, Taichung, Taiwan; 2. University of California, Irvine, CA

Address correspondence to J. Chen (jeonhc{at}uci.edu)

Objective: Fibrocystic change of breast, a common, noncancerous condition affecting more than 50% of women in their lifetimes, can sometimes simulate or make it difficult to detect a breast cancer with standard examination and imaging techniques. In mammography, the breast density associated with fibrocystic breasts may eclipse breast cancer on the mammogram film. Other imaging modalities might be needed. The purpose of this study is to analyze the MR imaging features of fibrocystic changes in the breast based on ACR BI-RADS MRI lexicon.

Materials and Methods: Thirty MRI of fibrocystic change of breast were found from 450 pathologically proved breast MRI in the past five years. Breast MRI was performed on a 1.5-T MR. Spin echo T1W sagittal and axial precontrast images and axial 3D SPGR dynamic contrast-enhanced images were acquired. Morphological and kinetic features were analyzed according to the ACR BI-RADS MRI lexicon. Types of lesions included focus/foci, mass and nonmass-like pattern. Enhancement kinetic curve was based on initial (within the first 2 minutes) and late phases. The initial enhancement phase is categorized into fast, medium, and slow. The delayed phase is described as persistent, plateau, and washout.

Results: Fifteen patients (15/30, 50%) presented as nonmass lesion, including twelve patch enhancement, two linear enhancement and one diffuse heterogeneous enhancement. Five patients (5/30, 17%) showed mass lesion with size smaller than 2 cm. Four patients showed focus lesion (4 mm). In six patients, breast MRI didn't show definite evidence of abnormal enhancement in the affected breast. Kinetic curve was created in sixteen patients including ten nonmass lesions and six mass or focus lesions. Eight nonmass lesions showed medium up-slope and persistent enhancement. Two nonmass lesions and all six mass or focus lesions showed rapid up-slope followed by wash-out (n = 4) or reaching plateau (n = 4) mimicking a malignancy.

Conclusion: The majority (50%) of fibrocystic changes in the breast presented as nonmass lesions on MRI. Kinetic enhancement curve can be either a benign or malignant feature, rendering it to be nonspecific. However, when a nonmass patch enhancement pattern, combined with medium up-slope kinetic enhancement curve, was found in breast MRI, fibrocystic disease of the breast is highly suggested and therefore should be included in the diagnostic list.

E005. Benign Breast Lesions Mimicking Carcinoma at Mammography: All You Need to Know

Muttarak M.; Na Chiangmai W.; Pojchamarnwiputh S.; Chaiwun B. Chiang Mai University, Chiang Mai, Thailand

Address correspondence to M. Muttarak (mmuttara{at}mail.med.cmu.ac.th)

Background: Many benign breast lesions pose diagnostic challenges. These lesions include abscess, hematoma, radial scar, postsurgical scar, diabetic mastopathy, fat necrosis, focal fibrosis, sclerosing adenosis, granular cell tumor, extraabdominal desmoid tumor, medial insertion of pectoralis muscle or sternalis muscle, and axillary lymphadenopathy (due to HIV infection, collagen vascular lesions, tuberculous and bacterial lymphadenitis). It is important for radiologists to be familiar with these benign lesions. Correlation of the patient's clinical features with the mammographic findings and additional use of sonography, fine-needle aspiration biopsy or core biopsy are helpful in establishing the final diagnosis and obviating unnecessary surgical intervention. In some of these lesions, surgery may be avoided while in others, the appropriate surgical procedure may be planned.

Key Issues: This exhibit aims to illustrate the mammographic features of carcinoma-mimicking benign lesions with pathologic correlation.

Format: Didactic. Imaging and pathologic correlation.

Teaching Points: 1. To describe the various causes of benign breast lesions that may mimic carcinoma on mammograms. 2. To highlight the important of knowledge of clinical presentation, additional use of sonography, fine-needle aspiration biopsy, and core biopsy in establishing the final diagnosis, in order to avoid unnecessary surgical intervention despite apparently-malignant mammographic features.

E006. The Many Faces of Fat Necrosis

Taboada J. L.¹; Whitman G.²; Krishnamurthy S.²; Stephens T.² 1. The University of Texas Health Science Center, Houston, TX; 2. The University of Texas M. D. Anderson Cancer Center, Houston, TX

Address correspondence to J. Taboada (Jorge.L.Taboada{at}uth.tmc.edu)

Background: Fat necrosis is a common entity which can pose challenges for the clinician and the breast imager. Presenting as a palpable mass or as an incidental finding on imaging, fat necrosis is usually due to an iatrogenic or a traumatic event. Pathologically, trauma can disrupt the fat cells of the breast resulting in a histiocytic response which leads to fibrous scarring. The pathologic response produces a spectrum of features seen on mammography, ultrasound, and magnetic resonance imaging (MRI).

Key Issues: This exhibit reviews the different presentations of fat necrosis. The appearance of fat necrosis ranges from typically benign to worrisome for malignancy. The various manifestations of fat necrosis are reviewed on mammography, sonography, and MRI, and correlated with histopathologic and cytologic findings.

Format: Didactic (Imaging findings with pathologic correlation)

Teaching Points: 1. To review the findings of fat necrosis seen on mammography, ultrasound, and MRI. 2. To correlate the imaging findings with cytologic and histopathologic findings.

E007. Granulomatous Mastitis: An Update of Clinical, Imaging, and Pathological Features Reported in a Large Series of 35 Cases

Hovanessian Larsen L.¹; Meiselman S.¹; Iyengar G.¹; Klipfel N.¹; Chantra P. K.²; Palmer S.¹ 1. University of Southern California, Keck School of Medicine, Los Angeles, CA; 2. VA Greater Los Angeles Healthcare System, Los Angeles, CA

Address correspondence to L. Hovanessian Larsen (lhovanes{at}usc.edu)

Background: Idiopathic granulomatous mastitis (IGM) is a rare, chronic inflammatory disease of the breast. This disease usually affects young women of child-bearing age or those with a history of oral contraceptive use. Since the clinical and radiological manifestations are similar to those of breast cancer, this condition is often initially misdiagnosed and proper treatment is delayed. Effective diagnostic protocols and treatment plans for IGM have not yet been established. Prior to 1980, surgical excision of the entire lesion was performed. More recently, treatment includes the use of corticosteroid therapy. We report on a large series of 35 IGM cases occurring in young women between the ages of 22 and 44 years, diagnosed at our center from January 2000 through September 2006. The purpose of this study is to review and describe the clinical, imaging, and pathological features of IGM and to discuss treatment alternatives in order to better diagnose and treat this rare disease.

Key Issues: This educational exhibit will review the imaging techniques used when evaluating IGM. Mammographic and sonographic features that are present at the time of initial diagnosis will be reviewed and correlated with clinical and pathological findings. We will describe the biopsy techniques, including FNA and core biopsy, most useful for diagnosis. We will present a protocol for follow-up with ultrasound imaging to document quantitative and qualitative changes occurring after treatment.

Format: This is a didactic presentation that will provide the viewer with the opportunity to examine the clinical, imaging, and pathological features of IGM. This presentation will include a discussion of biopsy techniques and recommendations for follow-up imaging of IGM. Current treatment options will also be discussed.

Teaching Points: After reviewing this exhibit the viewer will: 1. Describe the imaging features and differential diagnosis of IGM. 2. Identify the clinical and pathological features of IGM. 3. List the treatment and follow-up recommendations for the disease.

E008. Spectrum of Papillary Lesions of the Breast: Clinical, Imaging and Pathologic Correlation

Muttarak M.¹; Lerttumnongtum P.¹; Chaiwun B.¹; Peh W. C.² 1. Chiang Mai University, Chiang Mai, Thailand; 2. Singapore Health Services, Singapore

Address correspondence to M. Muttarak (mmuttara{at}mail.med.cmu.ac.th)

Background: Papillary lesions of the breast are a heterogeneous group of breast lesions that are difficult to diagnose as being benign or malignant. These lesions have varied morphologic features carry differing prognostic implications to affected patients, and the definitive treatment. Benign papillary lesions include papillary apocrine changes, papilloma, and papilloma with superimposed changes. Malignant papillary lesions include papillary ductal carcinoma-in-situ, micropapillary ductal carcinoma-in-situ, invasive papillary carcinoma, and invasive micropapillary carcinoma.

Key Issues: This exhibit illustrates the spectrum of clinical presentations, imaging features on different modalities, and pathologic correlation of papillary lesions of the breast.

Format: Didactic. Clinical, imaging, and pathologic correlation.

Teaching Points: 1. To review the clinical presentation, imaging features, and pathologic correlation of papillary lesions of the breast. 2. To facilitate the recognition of these lesions by appropriate additional imaging workup. 3. To discuss the prognosis and therapeutic interest of these lesions.

E009. Unusual Breast Lesions: Mammographic and Ultrasonographic Findings

Lee J.; Cho J.; Yoon S.; Oh J.; Choi J.; Nam K.; Cho S. Dong-A Medical Center, Busan, South Korea

Address correspondence to J. Lee (jhrad{at}dau.ac.kr)

Background: 1.To demonstrate the imaging findings in uncommon breast lesions 2.To become familiar with the imaging findings of a variety of rare breast lesions and consider them in the differential diagnosis although they are uncommon.

Key Issues: We illustrate the radiologic (mammographic and ultrasonographic) and the histopathologic features of unusual breast lesions that were confirmed pathologically in our hospital in the past three years. 1. Inflammation–tuberculosis. 2. Benign tumors–adenomyoepithelioma, fibromatosis (extraabdominal desmoid tumor), hamartoma, pleomorphic adenoma. 3. Malignant tumors–adenoid cystic carcinoma, infiltrating ductal carcinoma with osseous metaplasia, metastasis, fibrosarcoma in a male breast. 4. Miscellaneous–lymphangiectasia, hypertrophic scar after mammotome, Mondor's disease.

Format: 1. Format-didactic. 2. Organizational structure–we illustrate the mammographic and the ultrasonographic findings and provide radiologic-pathologic correlation.

Teaching Points: 1. A wide spectrum of uncommon benign and malignant lesions may be encountered at mammography and ultrasonography. 2. Even though they are rare, familiarity with these radiologic features will be helpful to diagnose various breast diseases.

E010. Pleomorphic Adenoma of the Breast: Case Reports and Review

Gonda S.; McGuire C. W. University of Kansas-Wichita, Wichita, KS

Address correspondence to S. Gonda (sgonda{at}kumc.edu)

Background: Over the past 100 years a variety of salivary gland-like tumors have been identified within the breast. One such tumor is pleomorphic adenoma. While this is the most common tumor of the parotid gland, it is rarely seen in the breast, with only 68 cases reported in the literature. Breast pleomorphic adenoma often has a typical gross pathologic appearance and is frequently located in the subareolar region. Scattered case reports and few reviews have been presented in journals of pathology and surgery, but little has been reported in the radiology literature. We present three cases, identified over the past eight years, of women who presented for biopsy of masses which were determined to be pleomorphic adenomas. Breast pleomorphic adenoma has a benign clinical course with no known deaths resulting directly from the tumor. However, histologic features can resemble those seen in malignancy, and because of this misdiagnosis often occurs. While no reported cases of malignant transformation have been identified in the breast, there have been examples in other soft tissues. Therefore, it is felt that soft tissue pleomorphic adenomas are at increased risk of malignant degeneration and thus these should be locally resected. Knowledge of the possibility of this tumor in the breast is therefore necessary in order to ensure appropriate treatment.

Key Issues: Breast pleomorphic adenoma is typically located in the subareolar region, most often seen as mass in the breast of a few centimeters in size. Must see myxoid or chondroid stroma histologically. Examples of mammographic, ultrasonographic and histologic findings will be shown.

Format: Didactic structure by imaging technique

Teaching Points: 1. Breast and salivary glands share morphological features and can also have similar pathologic processes. 2. Pleomorphic adenoma of breast is an example. 3. Only 68 cases reported, with women affected 10 times more than men. 4. The tumor is most frequently located in subareolar region. 5. DDx includes squamous carcinoma with prominent myxoid stroma, metaplastic carcinoma and phyllodes tumor. 6. The prognosis of breast pleomorphic adenoma is excellent. 7. While no evidence of malignant transformation in breast, there have been examples in other soft tissues and therefore local resection is recommended. 8. Misdiagnosis often occurs and has resulted in more aggressive surgical intervention than that which is indicated.

E011. Radiologic Manifestations and Histopathologic Correlation of Various Diseases Faced in Axilla

Park Y.²; Choi S.²; Lee S.²; Ryu J.²; Kim O.²; Lee J.³; Chu K.¹ 1. Busan National University Hospital, Busan, South Korea; 2. Busan Paik Hospital, Inje University, Busan, South Korea; 3. Dong-A University Hospital, Busan, South Korea

Address correspondence to Y. Park (pymrad{at}yahoo.co.kr)

Background: The purpose of this exhibit is to illustrate imaging findings of various interesting diseases arising in axilla and to correlate with pathologic findings.

Key Issues: Benign conditions include sarcoidosis, Kikuchi's disease, hemangiopericytoma, silicon-accumulated lymph node, fibroadenoma arising in accessory breast, abscess, tuberculous abscess, reactive hyperplasia caused by either Herpes zoster, or rheumatoid arthritis. Malignant diseases include malignant nerve sheath tumor, lymphoma, extraskeletal synovial sarcoma, infiltrating ductal carcinoma, metastatic lymphadenopathy from ovarian cancer, stomach cancer, malignant melanoma, hepatocellular carcinoma, or other various malignant diseases.

Format: For each case, imaging and histopathologic findings will be presented with the brief description of the disease including its course and management.

Teaching Points: It can be helpful in differential diagnosis and management of patients to be familiar with various axillary diseases.

E012. Uncommon Breast Cancer: Comprehensive Pictorial Review of Radiologic and Pathologic Findings

Woo O.; Yong H.; Kim A.; Park C.; Seol H.; Kang E. Korea University Guro Hospital, Seoul, South Korea

Address correspondence to O. Woo (wokhee{at}unitel.co.kr)

Background: Invasive ductal carcinoma and invasive lobular carcinoma of the breast are most common pathologic subtype of breast cancer, accounting for nearly 90% of all breast cancers. Uncommon breast cancer including not all types of breast cancer originate in a duct or lobule. The overlap of radiologic findings between common pathologic subtypes of breast cancer and uncommon breast cancers makes differentiation. However, some uncommon breast cancers have characteristic radiologic features that may suggest specific diagnosis. This exhibit illustrates the imaging findings of 40 uncommon breast cancers of the breast and axilla with pathologic confirmation.

Key Issues: To describe mammography, ultrasonography or MR imaging of uncommon breast cancer (apocrine carcinoma (n = 8), papillary carcinoma (n = 5), mucinous carcinoma (n = 4), medullary carcinoma (n = 4), tubular carcinoma (n = 4), metaplastic carcinoma (n = 3), malignant phyllodes tumor (n = 3), primary neuroendocrine carcinoma (n = 3), primary angiosarcoma (n = 2), primary breast lymphoma (n = 2), adenoid cystic carcinoma (n = 1), malignant peripheral nerve sheath tumor (n = 1)) with pathologic correlation.

Format: In present study, uncommon breast cancer accounts for 19% of all invasive mammary carcinomas. The overlap of radiologic findings between uncommon breast cancers and common pathologic subtypes of breast cancer makes differentiation difficult. Sometimes they mimic other benign tumors or inflammation. Usually, they commonly present as a relatively well defined mass, hypoechoic mass on sonography and with or without microcalcifications on mammography. Some uncommon breast cancers have characteristic radiologic features that may suggest specific diagnosis. For example, mucinous carcinoma shows nonenhancing low density areas within the tumor on CT. Familiarity with these radiologic findings of uncommon breast cancers can help ensure correct diagnosis and proper management.

Teaching Points: To know the clinical and imaging findings of the uncommon breast cancer with pathologic correlation and to provide radiological points for narrowing down the differential diagnosis of uncommon breast cancer.

E013. Imaging Findings and Clinical Presentation of Primary and Secondary Angiosarcomas of the Breast

Magut M.; Glazebrook K.; Reynolds C. Mayo Clinic, Rochester, MN

Address correspondence to M. Magut (magut.maureen{at}mayo.edu)

Background: Angiosarcomas are rare malignant tumors that arise from endothelial cells lining vascular channels. The breast is among the most common sites for angiosarcoma, with 8% of all primary angiosarcomas arising in the breast. Mammary angiosarcoma represents 0.04% of all malignant breast tumors. Primary angiosarcoma of the breast occurs sporadically in women in their thirties and forties; secondary breast angiosarcoma develops in older women several years following breast conservation therapy and radiation.

Key Issues: Mammary angiosarcoma typically presents as a painless rapidly growing palpable mass. A bluish discoloration of the overlying skin is the most specific clinical sign and is seen in 17-35% of patients. Histologically, breast angiosarcoma consists of endothelial lined anastomosing vascular channels invading the breast parenchyma. Mammographic appearance is usually that of a large ill-defined mass without spiculations. On ultrasound, breast angiosarcomas are generally poorly marginated with variable echogenicity. MRI typically shows a mass with low signal intensity on T1-weighted images and high signal intensity on T2-weighted images with rapid arterial enhancement and rapid washout.

Format: The proposed exhibit will be presented in a didactic format. Six cases of angiosarcoma of the breast (four primary and two secondary) will be reviewed. Each case will be discussed separately, starting with the clinical presentation and past medical history where relevant, followed by the imaging findings including ultrasound, mammography and MRI appearance. CT and nuclear medicine images will be included in some of the cases. Finally, the findings at pathology, how each patient was managed and the outcomes will be discussed.

Teaching Points: • Clinical presentation and risk factors of angiosarcoma of the breast. • Key differences between breast angiosarcoma and primary adenocarcinoma of the breast. • Mammographic, sonographic and MRI imaging characteristics and pathologic correlation of primary and secondary angiosarcomas. • Management of breast angiosarcoma.

E014. Metaplastic Carcinoma of the Breast: Radiological and Histopathological Findings

Cho K. R.; Kim K. M.; Seo B. K.; Lee K. Y.; Kim Y. H.; Chung K. B. Korea University Anam Hospital, Seoul, South Korea

Address correspondence to K. Cho (koreahl{at}dreamwiz.com)

Objective: To investigate the radiological and histopathological findings of metaplastic carcinoma of the breast.

Materials and Methods: The retrospective review of our institutional database during the last 7 years revealed 8 patients with histopathologically proven metaplastic carcinoma and they were included in this study. Two breast radiologists retrospectively reviewed the clinical and radiological findings by consensus. Radiologic findings were analyzed using the ACR BI-RADS (MM = 7, US = 6, low dose breast CT = 2, and MR = 2) and the histopathological findings were correlated with radiologic features.

Results: The mean age was 43.4 years old (33-55). There was no patient with a family history of breast cancer; all had palpable breast masses. On MM, all 7 patients showed oval shaped hyperdense masses without associated microcalcifications or architectural distortion. 4 patients had not-circumscribed margin. On US, 2 showed irregular shape and 4 had not-circumscribed margin. Internal echoes were hypoechoic and half of them showed complex echoes. Overing skin invasion (n = 1) and axillary lymphadenopathies were seen (n = 2) on both. On CT and MR, masses were peripherally enhanced and central nonenhancing areas, suggesting cystic necrosis histopathologically, were demonstrated. Invasion to skin and pectoralis muscle was seen in one. The histopathological subtypes were squamous (n = 3), matrix-producing (n = 2), sarcomatoid (n = 2), and mixed (n = 1). All patients underwent axillary LND and metastases were found in 3.

Conclusion: Metaplastic carcinoma of the breast frequently depicts as not-circumscribed oval shaped mass without microcalcifications or architectural distortion on MM, complex echoic mass on US, and peripheral enhancement with central low attenuation, suggesting cystic necrosis, on CT and MRI.

E015. Bilateral Synchronous Breast Cancer: A Comparative Study for Imaging Characteristics, Method of Detection, and Staging

Kim M.; Kim E.; Oh K. Yonsei University College of Medicine, Seoul, South Korea

Address correspondence to M. Kim (mines{at}medimail.co.kr)

Objective: This study was evaluated to compare the clinical, radiologic findings and the pathologic staging between first primary cancer and contralateral cancer of bilateral synchronous breast cancers and to determine the role of mammography and sonography in the detection of bilateral synchronous breast cancer.

Materials and Methods: The files of all patients operated on for primary breast cancer in our institute during the period 2000-2006 were retrospectively reviewed. Sixty two cancers in 31 patients, of which their imaging findings were available, made up our study population. We regarded the mass that a patient complained of or was detected early as the first primary cancer and the mass that was additionally detected in the contralateral breast during breast evaluation as a contralateral second primary cancer. We compared the clinical, mammographic, sonographic and pathologic findings between first cancer and contralateral cancer, and evaluated the detection method in each other.

Results: First and contralateral cancer showed statistically significant difference (p < 0.05) in the detection method, mammographic presentation, BI-RADS final assessment, sonographic boundary and posterior echogenicity, tumor size, and breast cancer staging, but no difference in BI-RADS lexicon in sonographic findings including shape (0.161), orientation (p = 0.712), internal echogenicity (p = 0.227), margin (p = 0.200) and calcifications (p = 0.328).

Conclusion: Additional contralateral breast cancers in bilateral synchronous breast cancers are apt to be small and less palpable, and to have less suspicious imaging findings and less advanced cancer staging. Sonographic evaluation with mammography has contributed to the early detection of bilateral synchronous breast cancer.

E016. Optimal CT Parameters Using Multi Detector-row CT (MDCT) in Evaluation of the Breast

Seo B.^1,2; Cho K.^1,2; Cho P.^1,2; Lee J.^1,2; Je B.^1,2; Lee K.^1,2; Lee Y.^1,2; Min B.^1,2; Kim B.^1,2; Cha S.^1,2 1. Korea University School of Medicine, Ansan, South Korea; 2. Korea University School of Medicine, Seoul, South Korea

Address correspondence to B. Seo (seoboky{at}korea.ac.kr)

Objective: It has been reported that CT is useful for initial staging of breast cancer, detecting tumor recurrence after treatment or follow-up modality for chemotherapy. However, the use of CT has been limited as a diagnostic tool of breast disorders because of radiation hazard. Radiation dose is determined by kVp, mAs, and collimation. The purpose of this study was to evaluate various radiation doses according to the CT parameters and provide the optimal parameter for breast CT scanning.

Materials and Methods: CT scanning was performed at various levels of kVp (80, 120, and 140 kVp) and mAs (30, 50, 100, 150, and 200 mAs). We used 16 x 0.626 mm collimation. We measured radiation dose using a CT body dose phantom. The phantom was made of solid acrylic, was 15 cm thick and had diameter of 32 cm. The phantom had four holes and we located two holes in the top and the remaining two holes in the bottom. For measurement of radiation dose in the phantom, we used a calibrated pencil ionization chamber and an associated radiation monitor controller. We measured radiation dose in each hole.

Results: Radiation doses with the phantom were 0.58-5.18 mGy at 80 kVp, 2.00-19.44 mGy at 120 kVp, and 3.10-25.91 mGy at 140 kVp. Less than 6 mGy was obtained at 80 kVp with any mAs, 120 kVp with 30 or 50 mAs, and 140 kVp with 30 or 50 mAs. Radiation dose in the top portion within the CT gantry was higher than in the bottom portion at all levels of mAs and kVp.

Conclusion: We recommend less than 50 mAs and prone position for breast CT scanning to reduce radiation hazard.

E017. Cone Beam Breast CT—A Feasibility Study With Surgical Mastectomy Specimens

Yang W. T.; Shaw C. C.; Chen L.; Altunbas M. C.; Wang T.; Lai C.; Kappadath C.; Tu S.; Liu X.; Sahin A.; Hunt K.; Bedrosian I.; Whitman G. The University of Texas M. D. Anderson Cancer Center, Houston, TX

Address correspondence to W. Yang (wyang{at}di.mdacc.tmc.edu)

Objective: Mammography is an important tool in the screening, diagnosis, and management of breast cancers. However, the effectiveness of film-screen and digital mammography has been compromised by the overlapping of cancers with breast anatomy. Cone beam CT can provide true 3-dimensional breast images with isotropic resolution (140 µm or smaller) and no reconstruction artifacts. We have constructed a flat panel cone beam CT system for breast imaging.

Materials and Methods: The experimental system consists of a general radiography tube pointing at a 30 x 40 cm² a-Si/CsI flat panel detector (Paxscan 4030CB, Varian Medical Systems, Salt Lake City, UT). A motor driven rotation stage holds and rotates the specimen to simulate dedicated breast CT imaging in which the patient would lie on a table in the supine position with one breast drawn downward through an opening to allow the X-ray tube and the detector to rotate and scan the breast. The scans are performed at 50-80 kVp with a voxel size of 140 or 300 µm. The dose level is estimated to be 1.5-1.8 cGy at the isocenter, corresponding to 2.5-3 times the mean glandular dose limit (0.6 cGy) for two view mammograms of a 5 cm thick compressed breast.

Results: The mean scanning time is 12 seconds for low resolution (binning) mode, which is adequate for visualizing tissue structures, and 48 seconds for high resolution (nonbinning) mode, necessary for visualizing small calcifications. Artifacts encountered from metallic tumor markers are successfully removed or reduced by postprocessing techniques without compromising tissue detail. Structured noise is minimal due to the absence of overlapping. Breast anatomy is well resolved on all images as skin, adipose, and glandular regions. Image noise is visible but low compared to the tissue contrast. Clear visualization of 250 µm or larger calcifications is successfully demonstrated. The detection of cancers is based on morphological assessment of tissue structures, improved compared to mammography, due to lack of overlapping. Cancers are visualized as areas of architectural distortion; irregular, spiculated masses with associated microcalcifications; or irregular masses with overlying skin thickening.

Conclusion: Cone beam breast CT demonstrates exceptional tissue contrast and can potentially reduce examination time with comparable radiation dose and eliminate the need for compression and additional workup views in mammography.

E018. A Three-Dimensional Image Reconstruction Algorithm for Creating High Quality CT Mammograms Utilizing Equivalent Mammographic Dose: A Preliminary Report

Vu H. T.²; Hrejsa A.¹; Han K. S.² 1. Advocate Lutheran General Hospital, Park Ridge, IL; 2. Rosalind Franklin University of Medicine and Sciences, North Chicago, IL

Address correspondence to H. Vu (hoang.vu{at}rocketmail.com)

Objective: To reconstruct and create high quality 3D computed tomography mammograms (CTM) utilizing equivalent mammographic dose

Materials and Methods: Gammex, acrylic test (50% adipose and 50% glandular tissue) and custom designed silicon/plastic phantoms with visible tumor masses and simulated micro-calcifications were employed for comparing the image quality between X-ray mammography and CT mammography (CTM). Degradations of the quality image due to artifacts were identified and compared in both phantoms. Phantoms were placed on a motorized, turnable table. The images of the phantoms were taken every 10 degrees (up to 180 degrees) and every 20 degrees (to 360 degrees). Equivalent dose and surface dose were measured with a source calibrated ion chamber at each acquisition angle. The actual center of rotation (COR) was calculated based on the image points. Arithmetic Reconstruction Technique (ART) was used to reconstruct the images with two main approaches: Technique A: used the maximum number of images (600 views) for superior image quality with quicker processing but with a tradeoff of increasing radiation dosage Technique B: used the fewer images with more computer iterations (longer processing) with possible poorer image quality but with less radiation dosage.

Results: Single digital image of the commercial phantom showed simulated lesions. Surface dose was found to be 175 mR and 400 mR for visualization of 0.24 mm and 0.16 mm micro-calcifications, respectively. ART reconstructed a mammographic test phantom showing a pair of 300 microns diameter holes. Two different image reconstruction approaches were employed using various numbers of views and iterations. Based on the preliminary results, the fewer the views, the more apparent or pronounced were the image artifacts. High contrast simulated lesions were visible with undesirable streak artifacts accentuated with fewer views. Low contrast simulated lesions were visible with increased grain noise accentuated with fewer views. In addition, previous work with ART was shown for comparative studies.

Conclusion: The preliminary result of using ART is promising and critical to develop CTM which potentially allows the radiologists to screen for high and low contrast breast lesions with acceptable radiation dosage and without breast compressions. Further evaluation of numbers of views and iterations should be done to optimize reconstruction. Investigation with breast phantoms of different compositions should be conducted.

E019. Timed Efficiency of Digital and Film-Screen Screening Mammographic Interpretation

Haygood T.²; Whitman G.²; Wang J.²; Atkinson E. N.²; Lane D.²; Patel P.¹; Stephens T.² 1. SUNY Upstate Medical University, Syracuse, NY; 2. U.T. MD Anderson Cancer Center, Houston, TX

Address correspondence to T. Haygood (tamarahaygood{at}yahoo.com)

Objective: Our purpose was to investigate the relative efficiency of interpretation of film screen (FSM) and digital (DM) screening mammography.

Materials and Methods: Using a specially-designed spreadsheet program, trained observers timed four radiologists as they read screening mammograms and noted the number of views provided, diagnostic aids used, diagnosis, and any peculiarities such as the presence of implants or a computer malfunction. FSM were prehung on a dedicated viewer. DM were viewed on a Stentor (Brisbane, CA) workstation. Reports were entered by the radiologists using a MagView (Burtonsville, MD) system on a separate computer. Data were analyzed using the Wilcoxon rank-sum and chi square tests.

Results: Four readers have so far interpreted 203 DM and 79 FSM. Excluding unilateral mammograms and cases with implants, 48 (25%) of DM consisted of the standard 4 views compared with 65 (82%) of FSM (p < 0.0001). Further excluding 29 DM and 5 FSM with unusual identifiable causes of prolonged reading times, measured interpretation speed for 44 4-view DM, 123 DM with more than 4 views, 57 4-view FSM, and 14 FSM with more than four views form the basis of this report. Mean measured interpretation speed for 4-view FSM was 74 seconds (range 52–206) compared with 160 seconds (range 88–341) for 4-view DM (p < 0.0001). 4-view DM may be further compared with DM with more than 4 views (mean 226 seconds, range 89-655) (p < 0.0001). Mean measured interpretation time for FSM with more than 4 views was 88 seconds compared with 226 seconds for DM with more than four views (p < 0.0001).

Conclusion: In our practice, DM using the Stentor workstation takes twice as long to interpret as FSM, when comparing only four-view studies and nearly three times as long to interpret when comparing studies with more than four views. Owing to the large percentage of DM with more than 4 views, the overall interpretation speed for DM is weighted towards the greater time needed for DM with more than four views.

E020. The Role of MR Spectroscopy in Breast Cancer Detection

Do R.¹; Moy L.¹; Salibi N.²; Mercado C.¹; McGorty K.¹; Hecht E.¹; Ruff J.² 1. NYU Medical Center, New York, NY; 2. Siemens Medical Solutions, St. Louis, MO

Address correspondence to R. Do (dok01{at}med.nyu.edu)

Objective: Magnetic resonance (MR) imaging is used in the detection of breast cancer. To distinguish between benign and malignant masses, both the morphologic and dynamic contrast enhancement (DCE) pattern of the lesions are evaluated. Studies have shown that MR spectroscopy (MRS) can detect choline, found predominantly in breast cancers. The purpose of this study is to determine whether choline can help distinguish between benign and malignant lesions and alter our BI-RADS categorization of breast lesions as evaluated by MRI.

Materials and Methods: Dynamic contrast-enhanced fat suppressed T1-weighted images were acquired at 1.5 T for 57 patients and 3.0 T for 16 patients. MRS data were obtained using a 1H SVS protocol with a PRESS sequence which includes spectral lipid and water suppression. A spectroscopy voxel was placed on enhancing breast lesion(s) for each patient. Two radiologists analyzed all lesions based on the morphologic and kinetic features, and assigned a BI-RADS category. Kinetic curves (DCE patterns) were characterized as persistent (type I), plateau (II), or early washout (III). The MRS data was included in a second readout session, along with the morphologic and kinetic analysis, to determine if the spectra changed our assessment of the lesion.

Results: In 70 patients, 73 lesions were found at MR that were subsequently biopsied. 53 lesions were diagnosed as cancers, including 13 ductal carcinoma in situ (DCIS), 38 invasive ductal carcinomas (IDC), and 2 invasive lobular carcinomas (ILC). A majority of cancer demonstrated the presence of a choline peak by MR spectroscopy (66%), with a greater proportion of IDC having choline (79%) than DCIS (23%). A majority of cancers (72%) demonstrated either a benign or indeterminate enhancement curve, with only 28% having a suspicious DCE pattern. Twenty lesions were benign, with the majority (13) diagnosed as fibroadenomas. Only 1 of 20 benign lesions demonstrated the presence of choline by MRS. Of the 53 cancers, the majority were categorized as BI-RADS 4, 5, or 6 (92%). 4 lesions were initially assigned a BI-RADS 2 category; 3 of these 4 lesions demonstrated a choline peak by MR spectroscopy.

Conclusion: A choline peak is detectable by MR spectroscopy in a majority of breast cancer lesions. It is a specific marker for breast cancer, being absent in the majority of benign breast lesions. MR spectroscopy provides predictive information not available with morphologic and DCE criteria alone in the evaluation of breast lesions on MRI.

E021. MRI Differentiation of Benign and Malignant Breast Lesions: Test Your Skills of the ACR BI-RADS Lexicon

Rausch D. R. Mount Sinai Medical Center, New York, NY

Address correspondence to D. Rausch (danarausch{at}hotmail.com)

Background: Contrast-enhanced breast MR imaging can be a powerful tool in the breast imaging armamentarium. When applied to the appropriate clinical setting and performed with the necessary technical requirements, it is a highly sensitive and reasonably specific method of breast cancer detection. The previously published American College of Radiology breast MR imaging lexicon helps to standardize the interpretation and reporting of these useful examinations.

Key Issues: As the use of contrast-enhanced breast MR imaging increases, it is critical that the interpreting radiologist understand and properly apply the terminology introduced by the lexicon. This exhibit will review this process with specific case examples.

Format: The cases will be presented in a quiz format. Key descriptive terms in keeping with the ACR BI-RADS MRI Lexicon and differential diagnostic points will be highlighted in the discussion of each case. The list of cases includes benign lesions such as sclerosing adenosis, fibroadenoma and fibrocystic change; high-risk lesions such as LCIS, intraductal papilloma and radial scar; and malignant lesions such as intraductal carcinoma, infiltrating ductal carcinoma, infiltrating lobular carcinoma and lymphoma.

Teaching Points: 1. To demonstrate the patterns of benign and malignant lesion enhancement on contrast-enhanced breast MRI with illustrated case examples 2. To apply the morphologic criteria of the ACR BI-RADS MRI Lexicon to 12 histologically verified lesions 3. To enhance skills of breast MRI interpretation.

E022. Proton Density, T1, T2 Clustering Patterns of Normal and Pathologic Tissues of the Breast Using Quantitative Magnetic Resonance Imaging

Trivedi M.³; Jara H.²; Tkacz J.² 1. Boston University College of Engineering, Boston, MA; 2. Boston University Medical Center, Boston, MA; 3. Boston University School of Medicine, Boston, MA

Address correspondence to M. Trivedi (mitesht{at}bu.edu)

Objective: The purpose of this study is to illustrate the relationship of proton density (PD), T1, and T2 density plots of the human breast using quantitative magnetic resonance imaging (QMRI). The hope is to show a clustering relation among different tissue types that can be used in future analysis of breast imaging to improve specificity and decrease the implications of false positive results.

Materials and Methods: Patients undergoing breast MRI and healthy volunteers were informed of the study and consented by institutional review board (IRB) approved protocols. Each patient was scanned using a 1.5-T Philips Intera Scanner (Philips Medical Systems; Andover, MA) using a mixed turbo spin-echo (mixed-TSE) sequence. The DICOM images were anonymized using ImageJ (National Institute of Health; Bethesda, MD) and the DICOM Rewriter plug-in (Walter O'Dell, Ph.D.; University of Rochester). A MathCAD (Parametric Technology Corporation; Needham, MA) program was developed to convert the DICOM file to a DICOM-RAW format. A separate MathCAD program was used to extract PD, T1, and T2 values from these images using a black-box algorithm. Each voxel from the PD, T1 and T2 maps created were redistributed in three-space: T1, T2, PD. Clustering patterns were analyzed in each data set and displayed graphically.

Results: Our results with normal breast tissue show that the voxels predictably form two ellipsoid distributions: fat and glandular tissue. Selection of predominantly fatty tissue in the image space isolates to a similar ellipsoid space in PD vs T1 vs T2 space among different patients. Similarly, selection of glandular tissue isolates to a distinct ellipsoid in the PD vs T1 vs T2 space among different patients. Similar relationships between the patients can also be appreciated on the y-view images, although for clarity the initial graphs are presented in T1 vs PD.

Conclusion: The results show a clear pattern of tissue distribution in the normal human breast. Two clouds of voxel data coalesce: fatty and glandular tissue distributions. Segmentation of the glandular data cloud may provide recognizable patterns in breast disease. The addition of kinetic curves to the variables may increase specificity. An approach can be contrived to superimpose QMRI voxel points over image space and detect morphologically suspicious lesions in the breast.

E023. Utility of Breast MRI in the Clinical Management of Patients with Mammographic or Sonographic Abnormalities

Raghu M.^1,2; Englander B.^1,2 1. Pennsylvania Hospital of the University of Pennsylvania Health System, Philadelphia, PA; 2. Pennsylvania Hospital of the University of Pennsylvania Health System, Philadelphia, PA

Address correspondence to M. Raghu (Madhavi.Raghu{at}uphs.upenn.edu)

Objective: The wide availability of breast MRI has led to an increased number of referrals for the clinical management of abnormal findings on mammography and/or ultrasound. Clinical management generally involves one of three options: annual mammography, short term follow-up, or biopsy—categorized as ACR BI-RADS categories 2, 3 or 4, respectively. Our objective is to determine if the clinical management of patients is altered based upon MRI findings.

Materials and Methods: The study was conducted as a retrospective review of patients without a history of breast cancer and who were categorized as ACR BI-RADS Category 2. Patients with a history of breast cancer were excluded. Correlation with the initial BI-RADS category and the MRI determined BI-RADS category was performed.

Results: 25 (n = 25) patients between the ages of 25 and 75 were evaluated and were initially categorized as BI-RADS 2. When the initial BI-RADS category was correlated with the finial MRI determined BI-RADS category, two result categories were created—a "change in BI-RADS" subset and a "no change in BI-RADS" subset. 18/25 (72%) demonstrated no change in the final BI-RADS assessment following MRI. 1 (4%) was recommended for biopsy (ACR BI-RADS 4A) based upon MRI findings, while 6 (24%) upgraded to BI-RADS 3 for close interval follow-up.

Conclusion: Breast MRI subsequent to mammography for findings considered benign (ACR BI-RADS 2) can alter the management of patients while not necessarily improving detection of malignancy in a screening population of patients without a history of breast cancer or markedly increased risk. It is possible that breast MRI may be over-utilized in this population and may play no significant role in the overall clinical management of these patients

E024. Pictorial Review: MRI Appearances of Invasive Lobular Carcinoma of the Breast (ILC)

Sharma N.; Manuel D. D.; Dall B. J. MRI Department, Cookridge Hospital, Cookridge, Leeds, United Kingdom

Address correspondence to N. Sharma (nishas{at}doctors.org.uk)

Background: Invasive lobular carcinoma (ILC) presents a diagnostic challenge because of its often subtle presentation clinically and on imaging. MRI of the breast is a useful adjunct to mammography and ultrasound both in making the diagnosis and in delineating tumor extent. If ILC is diagnosed at core biopsy and a wide local excision is planned, we offer MRI as part of the preoperative assessment.

Key Issues: MORPHOLOGY: ILC is subdivided according to morphological features on MRI into masses and areas of enhancement. The mass lesions are seen on both the high resolution T2 precontrast images and the post contrast subtraction images. The areas of enhancement are only seen on the post contrast subtraction images. ILC presented as a mass in 74% of patients (46/62) and as an area of enhancement in 24% of patients (15/62). One case was not identified on MRI (MRI occult). ILC presenting as a mass is either focal or multifocal. Focal masses can be spiculate 26% or smooth 3%. Multifocal masses can be a focal mass with satellites 14%, multinodular disease 8%, dendritic disease 10%, mixed multifocal and dendritic 8% and multicentric disease 5%. ILC presenting as an area of enhancement is subdivided into subtle 8% or overt 16% disease which usually reflects the size of the lesion. ENHANCEMENT PATTERN: Breast cancers, benign tumors and normal breast tissue will enhance after intravenous injection of gadolinium. Cancers usually will take up contrast more quickly and also it will washout more quickly because the tumor vascularity is abnormal. The enhancement curves are described as persistent (i.e., benign/indeterminate); plateau (i.e., suspicious) or washout (i.e. malignant). In our audit 20% of known ILC had a persistent type curve, 33% suspicious and 47% malignant type curve. There was no obvious correlation with morphological appearance and the type of curve obtained.

Format: This is a retrospective audit of 62 patients with ILC in the form of a pictorial review. T2-weighted images and post contrast T1-weighted subtraction images demonstrate the different morphological patterns. The different types of enhancement curves are also illustrated.

Teaching Points: 1.This review illustrates the spectrum of morphology and enhancement patterns of ILC seen on MRI. 2. The morphology is variable and the enhancement profile can be deceptively benign. 3. The MRI scan should be interpreted in conjunction with the mammography and ultrasound to optimize its benefit.

E025. MRI of the Postsurgical Breast

Siddall K.; O'Connell A. University of Rochester Medical Center, Rochester, NY

Address correspondence to K. Siddall (kristina_siddall{at}urmc.rochester.edu)

Background: As breast MR imaging (MRI) becomes more widely available and accepted, familiarity with common appearances of the postsurgical breast is increasingly important for accurate interpretation by the radiologist.

Key Issues: This exhibit will review morphology, signal characteristics and enhancement patterns of both benign and malignant changes seen on MRI of the postprocedural breast. Cases of MRI after breast biopsy will be shown, including examples of seroma, fat necrosis, and positive surgical margins. Breast MRI after lumpectomy, mastectomy and breast reconstruction will also be shown, with focus on MRI anatomy, detection of residual tumor, and sites of possible recurrence. The ideal timing of postprocedural breast MRI will also be discussed.

Format: This quiz will present breast MRI images as unknowns, followed by explanations of postoperative tissue features and pertinent mammographic and sonographic correlation.

Teaching Points: 1. Contrast-enhanced MR imaging is an effective adjunctive modality for assessment of the postoperative breast. 2. Shape, location, signal intensity on T2-weighted imaging, and dynamic enhancement behavior are important features for distinguishing between benign and malignant processes. 3. Proper timing of MRI and other imaging is essential for evaluation of the breast following biopsy or surgery.

E026. Clinical and Ultrasound Correlation of Common and Uncommon Breast Lesions from Neonates to Adolescence

Da Costa D.²; d'Almeida M.²; Cure L.²; Restrepo R.¹; Esserman L.² 1. Miami Children's Hospital, Miami, FL; 2. Mount Sinai Medical Center, Miami Beach, FL

Address correspondence to D. Da Costa (dpcosta{at}bellsouth.net)

Background: Ultrasound (US) is the ideal imaging modality for the evaluation and characterization of breast lesions in children and adolescents. Most of breast pathology in this age group represents benign disease, such as cysts, fibroadenomas, abscesses, cellulitis, and asymmetric breast tissue. On the other hand, malignancy is extremely uncommon; most are due to metastatic disease secondary to rhabdomyosarcoma, leukemia, lymphoma, and neuroblastoma.

Key Issues: Embryology and anatomy of the developing breast will be discussed. Breast tissue development begins in utero and continues through puberty. It has been classified into five Tanner stages based on physical exam and characteristic US appearances. US plays an important role in the work-up of breast lesions in the symptomatic patient. Additionally, color Doppler US is useful to differentiate between cystic and solid lesions, as well as, infectious processes.

Format: The educational exhibit will be organized in a didactic format. The study population consisted of female patients between 0 and 20 years of age, which were further subdivided into the following age ranges: 0 to 2 years old, 2 to 5 years old, 5 to 12 years old and 12 to 20 years old. A retrospective review of the clinical history and sonographic findings from September 2003 to September 2006 was performed. Correlation of clinical presentation and US findings were obtained. The presentation will be organized as follows: normal breast anatomy, embryology, and development (Tanner stages), physiological breast pathology, inflammatory/infectious and miscellaneous benign lesions, benign neoplastic lesions and malignant tumors. The prevalence of the most common pathologies in each age group will be emphasized.

Teaching Points: 1. Identify the US features of normal breast development and its normal variations, as well as, the most common breast pathologies in children and adolescents. 2. Identify the most common pathologies by age, location, and symptoms.

E027. Breast Ultrasound Screening in Asian Women with Dense Mammograms

Sim S.; Gogna A. Singapore General Hospital, Singapore

Address correspondence to S. Sim (gdrssj{at}sgh.com.sg)

Objective: To assess the feasibility and performance of breast ultrasound screening in Asian women with dense mammograms.

Materials and Methods: Between September 2002 and November 2004, 141 asymptomatic women attending our center for mammography with no clinical findings, negative mammograms and breast tissue density BI-RADS category 3 and 4 were offered second tier screening for breast cancer using ultrasound. The ultrasound examination was performed by a skilled breast ultrasonographer and verified when needed by a radiologist subspecializing in breast imaging. The median time taken to perform the scan was 12 minutes. The outcome was classified using the ACR BI-RADS system. Confirmation of ultrasound results was obtained via histopathological correlation for BI-RADS categories 4 and 5, and subsequent ultrasound/mammography for BI-RADS categories 1–3. Approximately 34.4% of women were peri- or postmenopausal. Twenty two (15.6%) were taking hormone replacement therapy concurrently. Seven (5%) had previous history of breast carcinoma. For these women, the analysis was restricted to the contralateral breast. In addition, 24 women had a positive family history of breast carcinoma among first degree relatives.

Results: Follow up data was available for 46.1% of the 141 women examined with ultrasound. Average length of follow up was 19 months. Biopsies were recommended in 9.2%. 2 cancers were detected solely by ultrasound, giving a cancer detection rate of 1.42%.

Conclusion: Breast ultrasound screening in the subset of Asian women with dense mammograms is feasible with a cancer detection rate of 1.42%. This compares favorably with that achieved with mammographic screening and other breast ultrasound screening trials performed in Caucasian women.

E028. Palpable and Nonpalpable Breast Cancers: Comparision of Ultrasonographic and Pathologic Findings

Cho K. R.; Seo B. K.; Choi E. J.; Lee K. Y.; Kim Y. H.; Kim J. H.; Chung K. B. Korea University Anam Hospital, Seoul, South Korea

Address correspondence to K. Cho (koreahl{at}dreamwiz.com)

Objective: To investigate any difference of ultrasonographic and pathologic findings between palpable and nonpalpable breast cancers.

Materials and Methods: 123 breast masses in 118 patients were included in this study, conducted from January 2001 to May 2003, in which surgically proven breast carcinoma (mastectomy 70, conservation therapy 53). The patients whose preoperative complaint was a palpable mass were categorized as group I (n = 73) and nonpalpable mass as group II (n = 50). We reviewed ultrasonographic and pathologic findings and investigated the difference between these two groups.

Results: On ultrasonography, spiculated margin was common in group I, 34% and ill defined margin was common in group II, 48% and irregular and oval shapes were in 71%, 11% in group I and 50%, 32% in group II (p < 0.05). Heterogeneous internal hypoechogenicity was 63% and 36% in each group (p < 0.05). There is no statistical difference in posterior echo pattern and the presence of internal microcalcification. According to BI-RADS category, 70% of group I was classified as category 5 and 60% of group II was category 4. On pathologic findings, infiltrating carcinoma was 86% and 72% in each groups and DCIS was 14% and 28%, respectively (p < 0.05). The mean lesional size was 27.1 mm and 20.7 mm. Axillary lymph node metastasis was seen in 50% and 19% in each two groups (p < 0.05). Under the TMN system, stage 0, 1, 2 was 12%, 12% 58% of group I and 28%, 36%, 34% of group II (p < 0.05).

Conclusion: There were statistically significant differences in ultrasonographic and pathologic findings between palpable and nonpalpable breast cancers.

E029. Ultrasonographic Findings of Papillary Lesions of the Breast: Differentiation between Benign and Malignant Tumors

Cho K. R.; Seo B. K.; Choi E. J.; Lee K. Y.; Kim Y. H.; Kim J. H.; Chung K. B. Diagnostic Radiology, Korea University Hospital, Seoul, South Korea

Address correspondence to K. Cho (koreahl{at}dreamwiz.com)

Objective: To understand the ultrasonographic findings of benign and malignant papillary tumors of the breast.

Materials and Methods: From July 1999 to June 2004, 81 lesions in 72 patients with surgically proven papillary tumors of the breast which demonstrated abnormalities on preoperative ultrasonographies (US) were included in this study. The pathologic results were benign in 65 cases (papilloma = 54, papillomatosis = 10, atypical papilloma = 1) and malignant in 16 (papillary ductal carcinoma in situ = 3, intraductal papillary carcinoma = 4, invasive PC = 9). 62 patients underwent surgery and 10 were US-guided core biopsy using 14G needle. One breast radiologist reviewed the clinical findings and two reviewed the US findings by consensus. US findings were analyzed by relationship between the mass and the duct. If the mass showed no relationship with duct, the mass was analyzed according to BI-RADS lexicon.

Results: The chief complaints of patients with benign tumor were for screening (n = 41), bloody nipple discharge (n = 16), palpable mass (n = 5), and mass with nipple discharge (n = 3). In malignant group, the number of each chief complaint was 4, 3, 7, 2, respectively. The US lesion size was 4–43 mm (mean, 9.3) in benign and that in malignant was 14-70mm (mean, 27.8). In benign papillary tumors, 47 lesions (72%) were related to the duct: Intraductal nodule (n = 17), intracystic nodule (n = 9), nodule with peripheral anechoic rim (n = 8), and nodule with adjacent ductectasia (n = 13). 18 cases were solid nodules without relationship to the duct and these were circumscribed isoechoic (n = 13), ill defined hypoechoic (n = 5. Central ducts were more frequently involved (70%). In malignant tumors, only 7 lesions (44%) were related to the duct: Intraductal nodule (n = 3), intracystic nodule (n = 1), and nodule with adjacent ductectasia (n = 3). Among the 3 cases of intraductal nodules, 2 were involved multiple ducts. The remained 9 cases (56%) were solid nodules without relationship to the duct: circumscribed oval (n = 1), irregular shaped (n = 2), and ill defined irregular (n = 6). Peripheral duct involvement is frequent (n = 4.57%). Among 10 axillary dissections, positive was only 2 and all of the invasive PC showed nuclear grade 1.

Conclusion: Malignant papillary tumors of the breast are less frequently related to the duct, more frequently involve peripheral duct, and often depicted as an irregular heterogeneous solid mass like other malignancies and larger in size on ultrasonography.

E030. Retrospective Breast Lesion Diagnosis Utilizing a Novel Ultrasound Strain Imaging Conspicuity Model

Wright S. L.²; Hesley G. K.²; Hangiandreou N. J.²; Morton M. J.²; Meixner D. D.²; Tzeng Y.¹; Lucas J.²; Nordland M.² 1. Cheng-Hsin General Hospital, Taipei, Taiwan; 2. Mayo Clinic, Rochester, MN

Address correspondence to S. Wright (wright.sonia{at}mayo.edu)

Objective: Ultrasound strain imaging (USSI) has shown promise in differentiating benign from malignant lesions based upon the lesion's innate firmness or compressibility. Specific USSI diagnostic criteria were developed and demonstrated great promise for characterizing breast lesions, when applied by the imaging technologist (tech). This study examines whether these criteria can be used retrospectively in the reading room.

Materials and Methods: 29 candidates were recruited from the clinical pool of patients scheduled for solid mass breast. Ultrasound (US) strain images were obtained prior to biopsy using a real-time strain imaging algorithm on a standard clinical Siemens Elegra US scanner. The tech recorded her "in-room" diagnosis at the time of scanning based on USSI data. Six retrospective evaluators viewed an instructional presentation on USSI diagnostic criteria. These "conspicuity" criteria included overall visibility of the lesion throughout the entire strain image sequence; good early visibility of the lesion, and visibility in contiguous sets of images; good lesion contrast; homogeneous lesion appearance; and increased lesion size on USSI compared to standard B-mode US. The evaluators consisted of two board-certified radiologists specializing in breast imaging, one radiology fellow, one radiology resident with basic breast imaging experience, and two experienced breast imaging techs (one of which acquired the USSI data). The evaluators first reviewed the available standard mammograms and US images for each patient and recorded likelihood of malignancy based upon a six point scale. Evaluators then viewed the strain sequence and recorded a second diagnosis. ROC analysis was used to assess overall diagnostic performance.

Results: Retrospective performance of the radiologists using USSI (NPV = 100% and 88%; specificity = 64% and 64%) was best in the evaluator group, and was similar to that of the "in-room" tech (NPV = 90%, specificity = 82%). Fellow and resident performance improved with USSI. The radiologist performance did not improve with USSI. Retrospective imaging tech results with and without USSI were mixed. Performance differences with and without USSI were not statistically significant (0.23 < p < 0.69).

Conclusion: Retrospective evaluation of USSI appears feasible. When the conspicuity criteria are applied retrospectively, USSI data added to the diagnostic value for the fellow and resident, but not for the radiologists in this study. This design should serve as a useful model for a future larger prospective study.

E031. Sonographic Patterns of Infiltrating Lobular Carcinoma (ILC) of the Breast

Whang I.; Lee J.; Kang B.; Cha E.; Park C. Cathollica University of Korea, College of Medicine, Diagnostic Radiology, Seoul, South Korea

Address correspondence to I. Whang (tiger{at}catholic.ac.kr)

Objective: To define the applicable US features of ILC for a clinically and mammographically challenging patient

Materials and Methods: We retrospectively reviewed pathologic confirmed breast cancer cases from our 4 branch hospitals. We built up a collection of 58 ILC cases: all mixed and not-mixed histology types were included. Of 58 cases, inappropriate low MHz probe used data plus only mammogram studied 16 cases were excluded. The remaining 42 cases using proper high MHz probe (1014 MHz) were taken into account to evaluate more exactly. The tumor was evaluated within the framework of evidence of mass, shape, margin, boundary, orientation, internal echogenicity, and shadowing based on ACR-BIRADS lexicon for breast US.

Results: In cases where there was not a definite mass, there existed architectural distortions or just variable posterior shadowings. The most common US pattern of our study was heterogeneous hypoechoic irregular shaped mass with variable shadowing (32/42). The following were only variable shadowing without discrete mass (6/42), architectural distortion (2/42), and mass presenting with sonographically benign or equivocal characters (2/42).

Conclusion: In conclusion, sonography is a helpful adjunct in the evaluation of ILC, a malignancy that often manifests as a subtle clinical and mammographic lesion. And in the case of sonography findings of shadowing without discrete mass or architectural distortion, these could be applicable characteristics in challenging cases.

E032. Ultrasonographic and Mammographic Findings of Nodular Sclerosing Adenosis

Park C.; Lee J.; Cha E.; Kim H.; Hwang-bo S.; Kang B. Department of Radiology, College of Medicine, The Catholic University of Korea, Incheon, South Korea

Address correspondence to C. Park (blounse{at}catholic.ac.kr)

Objective: To evaluate ultrasonographic and mammographic findings of cases of nodular sclerosing adenosis

Materials and Methods: We evaluated mammographic (n = 15) and ultrasonographic (n = 22) findings of 22 lesions with pathologically-proven nodular sclerosing adenosis in five hospitals for three years. We excluded adenosis cases combined with other breast disease. The review of mammography and ultrasonography was performed according to BI-RADS 4th edition.

Results: The age of the patients was 29–56 years (mean, 43.4 years). Twelve lesions were biopsied with percutaneous 14-gauge large core needle, 8 lesions with an 11-gauge vacuum-assisted mammotome needle and 2 lesions were removed with wire-localized surgical excision. The size of the lesions was 4-18 mm (mean, 0.8 mm). Ultrasonography (US) was performed in 22 patients, revealing 13 irregular, 8 oval and one round shape. The margin was circumscribed (n = 7) and noncircumscribed (n = 15; indistinct 5, microlobulated 5, angular 5). The orientation of mass was parallel in 13 cases and not-parallel in nine. The boundary was mostly abrupt interface (20/22). The echogenicity of mass was hypoechoic in 15, isoechogenicity in 4, hyperechogenicity in 2 and complex echogenicity in one. Nineteen cases showed no posterior feature. The vascular signals in the mass showed 4 cases on the color Doppler image. The final categories by BI-RADS were category 3 in 3 cases and category 4 in 19. Mammography was performed in 15 patients and distinct abnormalities were seen in 7. These consisted of an indistinct irregular mass in one having fatty breast composition, circumscribed masses in 3 and focal asymmetries in 3 patients having dense breast composition.

Conclusion: Nodular sclerosing adenosis mostly presents as category 4 on ultrasonography, therefore, biopsy is necessary to rule out malignancy. However, findings highly suggestive of malignancy such as the spiculated margin or echogenic halo are rare. On mammography, it is commonly obscured by dense breast.

E033. Sonographic Evaluation of Internal Mammary Lymph Node in Patients with Breast Cancer

Kim K. S.; Kim H. H.; Shin H. J.; Yang H. R.; Sohn J. H. Department of Radiology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea

Address correspondence to K. Kim (bleuaqua{at}lycos.co.kr)

Objective: The internal mammary lymph node status is important and influences therapeutic plan and prognosis in breast cancer. The purpose of this study is to evaluate the value of sonography in detection of internal mammary lymph node metastasis in patients with breast cancer.

Materials and Methods: From April 2003 to September 2005, authors reviewed 1352 patients who were diagnosed with primary breast cancer. Of 1352 patients, 654 patients (age range, 23-80 years; mean age, 45.1) underwent a preoperative sonographic evaluation of internal mammary lymph nodes. Sonographic criteria of metastatic internal mammary LN are as follows; more than 0.5 cm in the shortest diameter or more enlarged than contralateral internal mammary LN. Six hundred forty four of 654 patients were treated with surgery. Ten patients had chemotherapy and radiation therapy without surgical treatment. The median follow-up was 15.4 months. Two hundred twelve had follow-up sonography, and 83 had FDG-PET imaging. Metastatic internal mammary lymph nodes were confirmed by fine needle aspiration (FNA) (n = 4), surgical excision (n = 5) or FDG-PET (n = 12).

Results: Twenty-one patients had confirmed internal mammary lymph node metastasis. The incidence of metastatic internal mammary lymph node was 3% (21/654). Of 21 patients, 14 patients were diagnosed at preoperative sonography. Sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of sonography were 67%, 99%, 98%, 70%, and 99%, respectively.

Conclusion: Sonographic evaluation of internal mammary lymph node is useful for detecting lymph node metastasis in patients with breast cancer.

E034. Immersion Technique for Specimen US after US-Guided Needle Localization in Nonpalpable Breast Lesions

Seo B.; Jed B.; Min B.; Jung C.; Such S.; Kim Y. Korea University School of Medicine, Ansan, South Korea

Address correspondence to B. Seo (seoboky{at}korea.ac.kr)

Background: Specimen ultrasound (US) is required to confirm whether a nonpalpable breast lesion is completely resected after US-guided needle localization. However, qualified US images cannot be obtained in the air when the lesion or the specimen is small. To overcome this problem, we assessed the feasibility of immersion technique for specimen US in nonpalpable breast lesions after US-guided needle localization.

Key Issues: 40 surgical specimens were obtained by surgical excision. The size of specimens was from 22 mm to 40 mm, and the lesions ranged from 4 to 15 mm. The specimens were placed in a plastic container filled with saline. Specimens were scanned transversely and longitudinally with a high-frequency linear transducer, 12 MHz. There was no proven residual mass follow-up US after surgery.

Format: The format: didactic. Organizational structure: by imaging technique.

Teaching Points: In this exhibit, we will illustrate method of immersion US, usefulness of immersion technique for specimen US in nonpalpable breast lesions after US-guided needle localization and correlation of specimen US with pathologic findings.

E035. Sonographic Features of Gynecomastia

Dialani V.; Mehta T.; Baum J. Beth Israel Deaconess Medical Center, Boston, MA

Address correspondence to V. Dialani (vdialani{at}bidmc.harvard.edu)

Objective: To identify sonographic features of gynecomastia.

Materials and Methods: A retrospective analysis was performed on all male patients with breast symptoms presenting for imaging over a 5 year period. Breast ultrasounds in 153 men were jointly reviewed by the investigators without knowledge of clinical history. Ultrasounds were assessed for 1) masses (not present, nodular, ill-defined, spiculated or flame shaped), 2) location of mass (retroareolar, other) 3) vascularity (absent, present or not done), 4) lesion axis (parallel or perpendicular to chest wall, or neither), 5) appearance of posterior tissues (enhancement, shadowing, neither), 6) Tissue echotexture (hypoechoic, hyperechoic, mixed), and 7) AP diameter at the nipple (defined as increased if more than 1cm).

Results: 153 men (age 18-97 yrs), presented with pain (n = 38), lump (n = 95), both (n = 27), or nipple discharge (n = 3), 66 had bilateral and 87 had unilateral symptoms. Final diagnosis based on clinical and imaging findings and/or histology was gynecomastia in 148 (97%), angiolipoma in 2 (1.2%), abscess in 1 (0.6%), hematoma in 1 (0.6%) and breast cancer in 1 (0.6%). The 5 patients without gynecomastia had atypical imaging findings leading to intervention and histologic diagnosis. 9/148 with gynecomastia also had biopsy. 214 ultrasounds were performed of symptomatic breasts in patients with final diagnosis of gynecomastia. These revealed 69 (32%) masses: 16 nodular, 20 ill-defined, 33 spiculated or flame shaped. Most of the masses were retroareolar (n = 69; 100%), hypoechoic (n = 57; 83%), avascular (n = 54; 78%), parallel to the chest wall (n = 60; 87%), and showed no enhancement or shadowing (n = 47; 68%). Of the 145/214 (68%) without masses, 143 (99%) had increased AP diameter with hypoechoic (n = 76; 53%) or mixed (n = 61; 43%) tissue echotexture.

Conclusion: Gynecomastia is the most common cause of breast symptoms in men. In patients with high clinical suspicion of gynecomastia an ultrasound should suffice to make a diagnosis, if the described patterns are recognized.

E036. The Finding of Mucin on Breast Biopsy. How Should it be Managed?

Oneto J.; Vinocur D.; Esserman L.; Omarzai Y.; Poppiti R. Mount Sinai Medical Center, Miami Beach, FL

Address correspondence to J. Oneto (joneto{at}msmc.com)

Background: The presence of mucin in breast core biopsies is a finding commonly believed to be a harbinger of malignancy such as mucinous (colloid) carcinoma. However, there are benign entities that also produce mucin. Based on correlation with the histologic findings, radiologists are required to give clinical recommendations after performance of a breast biopsy that can range from follow-up examinations to surgical excision of the lesion. The pathologic and prognostic significance of finding mucopolysaccharide (mucin) in a biopsy varies according to its architectural distribution in the breast tissue as well as its pH (neutral or acidic). There are several pathologic criteria to determine the significance of the mucin. An understanding of these distinctions can avoid unnecessary workup. On imaging, the density and characteristics of the borders, as well as the presence of vascularity, and cystic or solid components gives a pattern that has been described to be related to the amount of mucin present.

Key Issues: A discussion with presentation of mammographic, sonographic and MR imaging of lesions containing mucin will be given. A complete analysis of the distribution of mucin, positivity in different stains and the pH, but mainly the distinction of the presence of mucopolysaccharide sometimes also called mucin, will be presented since it should be understood prior to making further management decisions.

Format: This educational exhibit will provide a comprehensive review in an interactive format of the mammographic, sonographic and MRI findings of mucin containing lesions. Pathologic correlation will be performed with a closer analysis of the pathologic findings to help the radiologist know how to react when the finding of mucin is reported in a breast core biopsy.

Teaching Points: To correlate and evaluate the imaging findings in patients with mucin producing lesions on biopsy. To analyze the pathologic findings and understand the terminology used in describing these lesions on core biopsy. To present a management protocol useful for radiologists when mucin is present in breast core biopsies.

E037. Magnetic Resonance Image-guided Breast Biopsy Equipment in a 3.0-T Environment: Safety and Image Artifact Evaluations

Kappadath S. C.; Stafford R. J.; Elliott A. M.; Whitman G. J. UT MD Anderson Cancer Center, Houston, TX

Address correspondence to S. Kappadath (skappadath{at}mdanderson.org)

Objective: To perform a preclinical magnetic resonance (MR) safety and artifact evaluation of interventional devices developed by Invivo (Orlando, FL) for MR-guided breast biopsies. Recent availability of interventional MR breast coils for 3.0-Tesla (3.0-T) fields necessitates evaluation of devices for interventional breast procedures in a 3.0-T environment.

Materials and Methods: The devices tested consisted of three metallic needles, three Teflon mandrins with Peek-polymer cannula, and a delivery tool for metallic clips. Device testing was performed in a clinical 3.0-T MR scanner (Excite HD, GE Healthcare, Milwaukee, WI). Image artifact evaluation was performed in a 7-channel breast array coil (GE Healthcare). The devices were positioned in an agar gel phantom via the biopsy localization plates built into the coil and imaged using 2D axial T1-weighted spin-echo (2DAxT1) and 3D fast-spoiled gradient-echo (3DfSPGR) pulse sequences, similar to those to be used for clinical protocols, to evaluate the extent of image artifacts in the 3.0-T MR environment. The safety of the devices was evaluated via measurements of static magnetic field-induced movement (deflection and torque) and measurements of the local radiofrequency heating measured by fluoroptic thermometry (m3300, Luxtron, Santa Clara, CA).

Results: None of the devices displayed any measurable magnetic field-induced torque or deflection (<3°) in a 3.0-T environment. No significant change in temperature (<0.5 °C) was measured in the phantom during 12 minutes of MR imaging at a specific absorption rate of 2 W/kg. The presence of Teflon devices introduced MR image signal distortions that extended 0.4–0.8 mm and ¬1–1.7 mm radially beyond their location for the 2DAxT1 and the 3DfSPGR pulse sequences, respectively. The presence of metallic devices introduced MR image signal distortions that extended 1.3–2.3 mm and 2.5–3.7 mm radially beyond their location for the 2DAxT1 and the 3DfSPGR pulse sequences, respectively. The tips of the Teflon and the metallic devices were easily visualized in the MR images; however, the metallic devices introduced asymmetric artifacts in the sagittal images with the 3DfSPGR pulse sequence.

Conclusion: The preliminary indications are that these devices may be used safely in an MR environment with static fields of 3.0 T or lower. The 2DAxT1 pulse sequence introduced smaller MR image artifacts than the 3DfSPGR pulse sequence. The MR image artifacts introduced by the Teflon devices were relatively minor.

E038. 1.5-T and 3.0-T Magnetic Resonance Imaging Artifacts from Breast Biopsy Clips

Le-Petross H.; Carkaci S.; Stafford R.; Elliott A.; Jackson E. MD Anderson, Houston, TX

Address correspondence to H. Le-Petross (huong.le-petross{at}di.mdacc.tmc.edu)

Objective: To evaluate the susceptibility artifacts of seven commonly used breast biopsy clip markers as seen on MR images obtained at 3.0 T compared to 1.5 T. A large susceptibility artifact from a biopsy clip can impair detection of small residual disease at the biopsy site or cause misinterpretation of the adjacent high signal artifact as residual disease, and larger susceptibility artifacts are expected at 3.0 T relative to 1.5 T unless appropriate acquisition parameter modifications are made. On the other hand, a biopsy clip with negligible artifact may not be visible. Therefore, when a small lesion is completely removed during a biopsy, the biopsy clip denoting the biopsy site may not be identified.

Materials and Methods: Seven metallic marker clips were imaged with 1.5-T and 3.0-T Excite MR Systems using dedicated phased-array breast coils (GE Healthcare Technologies, Waukesha, WI). The clips imaged include Micro-mark II, TriMark, Ultraclip II, BiomarC, Gel Mark, Gel Mark Cel-U-Gel, and SenoMark. Each marker clip was placed in the center of two custom-made Bovine gel phantoms. The clips were imaged using T2-weighted and T1-weighted gradient echo sequences with parameters identical to the clinical sequences used for breast MRI exams. The average size of the susceptibility artifact from each biopsy clip was measured.

Results: Susceptibility artifacts were visible on all MR sequences and with all seven breast biopsy clip markers. For the gradient echo sequence commonly used in dynamic series, the mean measurements were 7.1 mm at 1.5 T and 8.2 mm at 3.0 T. The difference was statistically significant (p = 0.0036). For T2-weighted sequence, there was no statistically significant difference between the artifact size for 1.5-T and 3.0-T images (p = 0.074). The average size of the artifact seen on T2-weighted sequences was generally more pronounced than on the gradient echo sequence, related primarily to the significantly higher bandwidth used in the gradient echo sequence.

Conclusion: The artifact size varied with the type and composition of the clip and the parameters of the MRI sequences used. The susceptibility artifacts were larger at 3.0 T than 1.5 T, but artifacts at 3.0 T could be controlled by appropriate increases in receiver bandwidth. It is important for the radiologists to be familiar with the artifacts from breast biopsy tissue markers since the size of the clip artifact could obscure pathology.

E039. Breast Lesions with Imaging-Histologic Discordance During US-guided 14G Automated Core Biopsy: Can the Directional Vacuum-Assisted Removal Replace the Surgical Excision? - Preliminary Report

Kim M.; Kim E.; Ki Keun K. Yonsei University College of Medicine, Seoul, South Korea

Address correspondence to M. Kim (mines{